Review - All FKBP Positives As well as Drawbacks

Twenty-five percent of all peaks were shared and showed high enrichments for terms related to gene regulation, such as ��regulation of transcription�� and ��transcription factor activity,�� suggesting that CLK/CYC are at the top of?a gene regulatory hierarchy across tissues. Indeed, 19.8% of the shared peaks lie close to TFs, a 5-fold SRT1720 supplier enrichment compared to all genes (p?= 2.36?�� 10?9). In contrast, we observed striking differences between the other sets: head-specific peaks were enriched in gene functions related to behavior and vision, while body-specific peaks were consistent with functions in metabolism (Figure?3B; Figure?S3 for full list). To test whether these differences were also reflected at the expression level, we first compared the CLK/CYC binding sites with gene expression according to FlyAtlas [25]. While shared CLK/CYC binding sites were close to genes expressed in head and body tissues, head- and body-specific peaks were located close to genes expressed either in heads or bodies (Figure?S4). Next, we compared CLK/CYC binding sites with data sets of cycling mRNAs [21]. We found genes close to head-specific peaks to be 3.4-fold enriched in genes cycling in heads (p?FKBP regulators in a TF hierarchy, CLK/CYC also bind directly to downstream targets in a cell type-specific manner, suggesting that they drive tissue-specific Pexidartinib purchase programs. The presence of E boxes alone cannot explain how, in each cell?type, CLK/CYC are recruited to distinct targets. As partner?TFs can help to define context-specific binding [3, 5, 6, 10?and?47], we searched for TF motifs that were differentially enriched between head and body binding sites. We found the ��orphan�� motif ME50 [48], opa, ME134/odd, and Adf1 to be?enriched in head-specific binding sites, while the motifs bab1, TATA/Mef2, ME3, GATA, and Hox were enriched in body-specific?binding sites (Figure?4A). To test whether combinations of differentially distributed motifs allow the discrimination of head versus body binding sites, we compared these sequences using a predictive binary classification framework [47]. Using sequence motif content alone, head and body peaks could be accurately distinguished using leave-one-out cross-validation (82% of peaks correctly classified; area under the receiver operating characteristic [ROC] curve [AUC]?= 0.91; Figures 4B and 4C). This indicates that partner TF motifs surrounding CLK/CYC binding sites carry information indicative of binding and have the potential to determine context-specific clock target genes and function. It further suggests that the corresponding TFs could be novel tissue-specific CLK/CYC partner factors.